Sftware as a Service

Discussion of anything and everything relating to chess playing software and machines.

Moderator: Ras

User avatar
Rolf
Posts: 6081
Joined: Fri Mar 10, 2006 11:14 pm
Location: Munster, Nuremberg, Princeton

Re: Sftware as a Service

Post by Rolf »

towforce wrote:
Matthias Gemuh wrote:If I buy software and it is not sitting in my computer, it has to be in my cupboard. Otherwise the developer has stolen my money.
When you go on holiday, presumably you buy a car after you arrive, and sell it again before you go. You couldn't allow your money to be stolen by renting one! :wink:
Hehe, good one.

I must admit that I was shocked when I read - not so many years ago, after I had used *my* chess software all the time - that they write in their ReadMe, that after the following conditions blabla, I were allowed to use the following software but it wer not my property. Something like this.

I thought, crazy, but this is just a legal reservation without any practical meaning.

Now I begin to understand that basically nothing has changed.

Look, the older versions of say Fritz are still in my cupboard, but I dont care at all. I dont want to use it anymore and I couldnt sell it either because who is so stupid to use outdated software?

Well, crazy enough, on Ebay they buy such stuff. Because for lays even older software plays chess well enough.

But now think of this. If that happens, then I have a cluster online and the most recent update of Rybka and use something that they only use at Wch. And I can have the output on that level.

Why should I fill my cupboards with the mere material when in chess it's all about thoughts and speed and electricity?

But sorry that I hijacked that thread because now certainly the whole forum will decide that online flexibility is crap and water-cooling a cluster at home is wisdom! :)
-Popper and Lakatos are good but I'm stuck on Leibowitz
User avatar
Dr.Wael Deeb
Posts: 9773
Joined: Wed Mar 08, 2006 8:44 pm
Location: Amman,Jordan

Re: Sftware as a Service

Post by Dr.Wael Deeb »

Rolf wrote:
towforce wrote:
Matthias Gemuh wrote:If I buy software and it is not sitting in my computer, it has to be in my cupboard. Otherwise the developer has stolen my money.
When you go on holiday, presumably you buy a car after you arrive, and sell it again before you go. You couldn't allow your money to be stolen by renting one! :wink:
Hehe, good one.

I must admit that I was shocked when I read - not so many years ago, after I had used *my* chess software all the time - that they write in their ReadMe, that after the following conditions blabla, I were allowed to use the following software but it wer not my property. Something like this.

I thought, crazy, but this is just a legal reservation without any practical meaning.

Now I begin to understand that basically nothing has changed.

Look, the older versions of say Fritz are still in my cupboard, but I dont care at all. I dont want to use it anymore and I couldnt sell it either because who is so stupid to use outdated software?

Well, crazy enough, on Ebay they buy such stuff. Because for lays even older software plays chess well enough.

But now think of this. If that happens, then I have a cluster online and the most recent update of Rybka and use something that they only use at Wch. And I can have the output on that level.

Why should I fill my cupboards with the mere material when in chess it's all about thoughts and speed and electricity?

But sorry that I hijacked that thread because now certainly the whole forum will decide that online flexibility is crap and water-cooling a cluster at home is wisdom! :)
Making hidden advertisement,eh uncle Rolf :!: :?:
Well,you said it yourself,it's a crap....
_No one can hit as hard as life.But it ain’t about how hard you can hit.It’s about how hard you can get hit and keep moving forward.How much you can take and keep moving forward….
bob
Posts: 20943
Joined: Mon Feb 27, 2006 7:30 pm
Location: Birmingham, AL

Re: Sftware as a Service

Post by bob »

michiguel wrote:
bob wrote:
CRoberson wrote:The idea has been around for close to a decade and I've thought about releasing software that way. It seems to be a reasonable idea
given that internet access is practically ubiquitous and becoming more
so. Also, it provides a solution for the software theft issues.

OTH, it is a paradigm shift that consumers will stumble over. Can it be better than what goes on now? Sure, why not? Isn't just a matter of
doing it right? Of course it is. The big question is how to do it right.

I never knew that Vas was thinking of it. Seems a bold step to leap where others haven't tried yet.

What made me start thinking about this model years ago? Simple: it was all the software theft that is going on. The current
model seems to be in trouble. Vendors raise their price due to all the theft and then consumers are pissed. Some claim
that leads to more theft, which leads to higher prices ....

Maybe a new model is needed. Shouldn't the developers have some protection from all the theft or are people more interested in all
the free clones that come from the current system?

Despite some ohter peoples claims, modern programs have algorithms and heuristics that weren't invented 30 years ago.
Sometimes we can use old ideas, but because it can be used. Sometimes old ideas get new twists that make them usable. So, its not just the old idea.
It works for some things. But not for chess. how do you "reach" the service when flying across the country and you want to do a bit of preparation for an upcoming tournament while you have some down time? Providing a service means that the end user has to also provide a service to make the connection, and this is not always practical, or affordable. And it negates the idea of having a computer along with you. Should everybody just buy netbooks and the like from now on? What about performance? Providing enough horsepower to make this work also looks both problematic and expensive. Seems like a solution to a problem that has yet to be identified.
OTOH, you reach the power of cluster with your cell phone. Different business models can coexist to satisfy different interests.

Miguel
Maybe. Or maybe you reach a cluster that is already heavily loaded, and offers _you_ computing power that is actually less than that which you have in your cell phone CPU. What then?
User avatar
Matthias Gemuh
Posts: 3245
Joined: Thu Mar 09, 2006 9:10 am

Re: Sftware as a Service

Post by Matthias Gemuh »

towforce wrote:
Matthias Gemuh wrote:If I buy software and it is not sitting in my computer, it has to be in my cupboard. Otherwise the developer has stolen my money.
When you go on holiday, presumably you buy a car after you arrive, and sell it again before you go. You couldn't allow your money to be stolen by renting one! :wink:
Well, it's about staying at home and paying for a car that remains at vendor's parking lot :D .

Matthias.
My engine was quite strong till I added knowledge to it.
http://www.chess.hylogic.de
gladius
Posts: 568
Joined: Tue Dec 12, 2006 10:10 am
Full name: Gary Linscott

Re: Sftware as a Service

Post by gladius »

bob wrote:Maybe. Or maybe you reach a cluster that is already heavily loaded, and offers _you_ computing power that is actually less than that which you have in your cell phone CPU. What then?
I don't really want to get into this argument, but a big part of "cloud computing" is dynamic scalability. It would be very expensive to connect to 512 cores, as you'd have to pay to run 512 nodes, but the backing infrastructure of things like Amazon EC2 and Microsoft Azure would allow you to scale the load dynamically.

If 15 people wanted to pay enough to do so, they could. Then once they are done, you can shut off the cores, and they go back to the general pool (and most importantly, you don't have to pay for them :)).
bob
Posts: 20943
Joined: Mon Feb 27, 2006 7:30 pm
Location: Birmingham, AL

Re: Sftware as a Service

Post by bob »

gladius wrote:
bob wrote:Maybe. Or maybe you reach a cluster that is already heavily loaded, and offers _you_ computing power that is actually less than that which you have in your cell phone CPU. What then?
I don't really want to get into this argument, but a big part of "cloud computing" is dynamic scalability. It would be very expensive to connect to 512 cores, as you'd have to pay to run 512 nodes, but the backing infrastructure of things like Amazon EC2 and Microsoft Azure would allow you to scale the load dynamically.

If 15 people wanted to pay enough to do so, they could. Then once they are done, you can shut off the cores, and they go back to the general pool (and most importantly, you don't have to pay for them :)).
I'm aware of the ideas behind cloud/grid computing and such. But I am also aware of the issues related to chess, which is ridiculously computationally intensive in nature. I suspect the market for chess is small enough (ignoring the chessmaster-type end users that could care less about high-speed hardware anyway) that it is simply an idea that won't work very well. We have such facilities here. In fact, I recall one example where the state bought a big Cray computer years ago, but they had enough users that the serious researchers decided to buy their own smaller machines, because a dedicated small machine was much faster than a tiny slice of a large super-computer. I don't believe that more than 100 people in the world would really want/need access to a large chess cluster, when they can already buy 8-24 core boxes at reasonable prices (4 x 6 core AMD boxes are not inexpensive, but they are not 20K boxes either). When your personal computer exceeds the instantaneous power you can realistically extract from a cluster, the cluster becomes less useful. And if you have absolutely no way to fall back and run on your faster local hardware, then you have a real problem.